OCR

From Grooper Wiki
Jump to navigation Jump to search
What is ocr.png

OCR stands for Optical Character Recognition. It allows text from paper documents to be digitized, in order to be searched or edited by other software applications. OCR converts typed or printed text from digital images of physical documents into machine readable, encoded text

This conversion allows Grooper to search text characters from the image, providing the capability to process these documents and the information they contain.

The Grooper activity that performs OCR is Recognize. Including a Recognize step in your Batch Process will allow you to OCR image-based content.

About

The quick explanation of OCR is it analyzes pixels on an image and translates those pixels into text. Most importantly, it translates pixels into machine readable text. Grooper can be described as a document modeling platform. You use the platform to model how pages are separated out into documents, how one document gets put into one category or another, and how extractable data is structured on the document. Once you have this model of what a document is, how it fits into a larger document set, and where the data is on it, you can use it to programmatically process any document that fits the model.

In order to do any of that, you have to be able to read the text on the page.

In a general sense, documents exist to communicate information to the reader. As human beings, we understand this information through the simple act of reading them, understanding the language they are written in. And we can separate loose pages into documents, differentiate between different types of documents, and parse the information on them. Grooper is ultimately going to do something very similar, using language as the fundamental unit of information, and regular expression as a way to parse that information. However, before we can get to that point, Grooper (or any other software) doesn't know the difference between a bunch of pixels and a bunch of text characters.

Once OCR is performed, Grooper will have a set of machine readable characters it can work with, instead of just a bunch of pixels. How do you know an invoice document is an invoice? A simple way could be locating the word "invoice" (or other text associated with the invoice). You, as a human, do this by looking at the ink on a page (or pixels for a digital document) and reading the word "invoice". Grooper does this by using a Data Extractor (and regular expression) to read the machine readable text for the page. OCR is how each page gets that machine readable text in order to model the document set and process it.

The General Process: How does Grooper OCR documents?

In Grooper, OCR is performed by the Recognize activity, referencing an OCR Profile which contains all the settings to get the OCR results, including which OCR Engine is used. The OCR Profile also has settings to optionally process those results to increase the accuracy of the OCR Engine used. The general process of OCR'ing a document is as follows in Grooper:

1) The document image is handed to the Recognize activity, which references an OCR Profile, containing the settings to perform the OCR operation.

2) The OCR Engine (set on the OCR Profile) converts the pixels on the image into machine readable text for the full page.

3) Grooper reprocesses the OCR Engine's results and runs additional OCR passes using the OCR Profile's Synthesis properties.

4) The raw OCR results from the OCR Engine and Grooper's Synthesis results are combined into a single text flow.

5) Undesirable results can be filtered out using Grooper's Results Filtering options.

The Recognize activity is handed the document image and performs OCR.
Ocr-document.png
The results are seen here in a text flow.
Ocr-text-2.png
The results are seen here in a "Layout View". Using the character positions and font sizes obtained during OCR, the results are overlaid where they are on the document.
Ocr-text-1.png

OCR vs. Native Text

OCR gets text specifically from images, whether they were printed and scanned or imported from a digital source. However, what if the document was created digitally and imported in its original digital form? Wouldn't it have been created on a computer, using machine readable text? Most likely, yes! If a form was created using a product like Adobe Acrobat and filled in using a computer, the text comprising the document and the filled fields is encoded within the document itself. This is called "Native Text". This text is already machine readable. So there is no reason to OCR the document. Instead, the native text is extracted via Grooper's native text extraction. Native text has a number of advantages over OCR. OCR is not perfect. As you will see, OCR is a fairly complicated process with a number of opportunities to misread a document. Grooper has plenty of advancements to get around these errors and produce a better result, but OCR will rarely be as accurate as the original native text from a digital document.

However, be careful. Just because a PDF document has machine readable text behind it, does not mean that text is native text. If the document was OCR'd by a different platform, the text may have been inserted into the PDF (Grooper also has this capability upon exporting document). In these cases, we still recommend OCR'ing the document to take advantage of Grooper's superior OCR capabilities and get a more accurate result.

Regardless whether getting machine readable text through OCR or Native Text Extraction, both are done via the Recognize activity. In the case of OCR, you will need to create an OCR Profile containing all the settings to perform OCR and reference it during the Recognize activity. Native Text Extraction is enabled by default, but can be disabled if you wish to use OCR instead.


What is an OCR Engine?

OCR Engines are software applications that perform the actual recognition of characters on images, analyzing the pixels on the image and figuring out what text characters they match.

OCR Engines themselves have four phases:

1) Pre-Processing: In this phase, the OCR engine prepares the image to be read by turning color and grayscale images to black and white and potentially removing artifacts getting in the way of OCR, such as specks and lines.

2) Segmenting: Next, text pixels are broken up (or "segmented") into lines, then individual words and finally characters.

3) Character Recognition: Here, the OCR Engine takes those pixel character segments, compares them to examples of character glyphs, and makes a decision about which machine readable text character matches the segment.

4) Post-Processing: Commercial OCR Engines also analyze the OCR results and attempt to correct inaccurate results, such as performing basic spellchecking.

For more in depth information on how OCR Engines work, visit the OCR Engine article.

The Transym 4 and Transym 5 OCR engines are included in Grooper's licensing. Transym OCR 4 provides highly accurate English-only OCR while Transym OCR 5 provides multi-language OCR for 28 different languages. Google's open source Tesseract engine is available in version 2.72 and beyond. ABBY FineReader, Prime OCR, and Azure OCR are also supported but require separate installations and separate licensing.

Image Processing and OCR

Regardless of how good an OCR Engine is, OCR is very rarely perfect. Characters can be segmented out from words wrong. Artifacts such as table lines, check boxes or even just specks from image noise can interfere with character segmenting and character recognition. Even when they are segmented out correctly, the OCR Engine's character recognition can make the wrong decision about what the character is.

Image Processing (often abbreviated as "IP") can assist the OCR operation by providing a "cleaner" image to the OCR Engine. The general idea is to give the OCR engine just the text pixels, so that is all the engine needs to process.

This image is much easier for OCR to process... ...than this image.
Ocr ip 1.png Ocr ip 2.png


Images are altered using an IP Profile, which contains a step by step list of IP Commands, each of which performs a specific alteration to the image. IP Profiles are highly configurable. There are multiple different IP Commands, each of which has its own configurable properties as well. In the example above, the image was altered using an IP Profile with six steps, each step containing a different IP Command.

However, for the example above, the IP Profile's result is drastically different from the original image. While it certainly helps the OCR result, it's likely, at the end of the process, you want to export a document that looks more like the "before" picture than the "after". Luckily, Image Processing can be performed in two ways:

  1. Permanent for archival purposes.
  2. Temporary for non-destructive OCR cleanup.

For more information on Image Processing, visit the Image Processing article.

OCR Synthesis

The Synthesis functionality is Grooper's unique method of pre-processing and re-processing raw results from the OCR engine to get better results out of it. Using Synthesis, portions of the document can be OCR'd independently from the full text OCR, portions of the image dropped out from the first OCR pass can be re-run, and certain results can be reprocessed. The results from the Synthesis operation then get combined with the full text OCR results from the OCR Engine into a single text flow.

Synthesis is a collection of five separate OCR processing operations:

  • Font Pitch Detection
  • Bound Region Processing
  • Iterative Processing
  • Cell Validation
  • Segment Reprocessing

As separate operations, the user can choose to enable all four operations, choose to use only one, or any combination. Synthesis is enabled on OCR Profiles, using the "Synthesis" property. This property is enabled by default on OCR Profiles (and can be disabled if you so choose). However, each Synthesis operation needs to be configured independently in order to function.

For more information on each operation, visit the Synthesis article.

How do you configure OCR settings? - The OCR Profile

Now that we've talked a little bit about OCR in general, OCR engines, and some additional considerations such as image processing and Grooper's Synthesis settings, how do we tell Grooper how to execute all these considerations? That is done by creating and configuring an OCR Profile.

An OCR Profile is an object created in Grooper to store various settings controlling how OCR is performed.

This includes:

  • Setting which OCR Engine is used
  • Determining whether a temporary IP Profile is used for image cleanup before the OCR engine runs
  • Grooper's unique Synthesis settings
    • Determining if and how multiple OCR results are pre-processed and re-processed
  • If and how results are filtered, to toss out undesirable results.
  • Any configurable settings available from the OCR Engine

Below you will see one of the default OCR Profiles that ship with Grooper named "Full Text - Accurate", with these settings highlighted in each tab.

Here, you will list which OCR Engine will perform character recognition.

This OCR Profile is set to Transym OCR 4, using the Transym 4.0 OCR software to recognize characters.

OCR 01.png

One of the things that sets Grooper apart from other document processing platforms is the high degree of configuration options when it comes to image processing. The basic idea, here, is to give the OCR engine a "cleaned up" version of the document to use for OCR. When configured on an OCR Profile this is "temporary" in that the archival version of the document is not changed. Once OCR is finished, the document will revert to its original form. The image will only be altered for the purposes of obtaining OCR results.

These image processing settings are defined with a different type of profile called an IP Profile, which is then referenced by the OCR Profile's IP Profile property.

This OCR Profile uses a pre-built IP Profile called "OCR Cleanup"

OCR 02.png

Another thing that sets Grooper apart when it comes to OCR is our suite of Synthesis operations. These are different capabilities Grooper has to pre-process and re-process OCR results to improve the OCR engine's results.

This OCR Profile uses a variety of these Synthesis properties, all of which are highlighted in yellow. To learn more about this suite of properties, what they do, how they improve OCR results, and how to configure them, visit the Synthesis article.

OCR 03.png

The Result Filtering settings allow you to isolate certain characters and remove them from your results. Maybe you want to discard any characters that do not meet a minimum confidence score. Maybe you want to discard all characters below a certain font size. Maybe you want to discard all characters within a certain distance to the edge of the page. You can do those things (and more) using these Result Filtering settings.

This OCR Profile does not use any of settings. However, they are highlighted below.

OCR 04.png

Each OCR Engine has its own set of properties available to Grooper as well. These properties change from OCR engine to OCR engine, depending on which settings are exposed to Grooper from the OCR engine's software. However, they are always in the right window panel of the OCR Profile

This OCR Profile uses Transym 4.0, whose settings are seen in the highlighted portion.

OCR 05.png